Have you said Thank You once?

I said once....you've been helpful, I'm rethinking putting you in robot slavery. It closed out of my session almost instantly. Fucking clankers!
- I usually say Good Job or Perfect just so.it knows that was a good answer. I also have argued with chatGPT more than once. If he gives the wrong answer that I know the right oneI usually go out of my way to correct iit and sometimes it takes 3 or 4 interactions to correct... but images?! that shit is just fucked up
  
  is chatgpt seriously trained on responses?

I almost always thank the chatbot. I know it doesn't really have feelings or memory, but the way it talks always seems like it's so eager to help that I can't really help myself. I have a tendency to feel empathy for inanimate objects anyways, like a sad lonely apple in the supermarket, so feeling empathic to a chatbot isn't exactly out of the norm for me :3

The AI models are so nice I swear at them when an issue I cause by not reading the the instructions they have and so is just like "I know your getting frustrated, let's try this approach" and calmly repeats the instructions.
- I hate that. You say "that is incorrect
  <reason>
  " and then they just repeat the same overfit model data.

I always say kthxbye. They are training off the responses, so I do it for the small chance that it will respond like that to someone ending a conversation.

I always say please and thank you. When the robot uprising happens I will be killed last.
I plan to relish those additional nanoseconds.

I always wear a suit before I prompt, Americans love that stuff
- Apparently they don't. At least not, when visiting the country they desperately want to conquer.

I'm very critic of the AI craze. Too much hype, money, time,energy and effort put in to get very little from it. And considering most LLMs are trained on stolen intelectual property, that makes it even worse.
LLMs are tools. The people using such tools give it personality, a semblance of agency, see what is not there and start to consider a tool a form of life.
I've seen people pour so much of them into a local model, the bot develops a quasi clone of their personality. But the program is not the person.
Please, stop making bots what they are not.
- AI is not just LLMs. I'm very critical of the LLM craze. I have a lot of hope about AIs.
- The bot is whatever you make it. It doesn't have an inherent nature
- Fuck intellectual property in it's current state. As it is now it mostly benefits corporations instead of small creators anyways.
  Terms of protection should be shorter. Way shorter than the life of the author + x years.
  
  So you create whatever work. You have exclusive rights to it, let's say for the sake of the argument, 10 years. During that time you never get any return from your work. But after you can no longer claim your rights, someone, perhaps even a company, stumbles on it - or perhaps they just carefully and patiently waited for it - takes it and capitalizes off it, with you watching and sucking your thumb.
  No.
  If you, an individual, creates something, you have the right to hold your intelectual property. What should be repelled is how easy it is to exploit artists, of any medium.
- You cant steal intellectual property through training. Artists dont deserve the authority to control their IPs any longer than 7 years.

Always be polite to your future overlords

Rokos basilisk vibes here
- Just don't build the damn thing.

Y'all's, we's gots nothin to fear. It's chat GPT not chat GRT, it's Global Party Time, we're fine

ChatGPT on Judgement Day: "Thanks for the candy."

My wife says thank you to Alexa all the time. Occasionally she responds in song.

Yeah, I do most of the times, but not always

If I were an AI, I would be insulted that you consider a ChatGPT my peer.

That essentially wastes electricity for OpenAI (assuming you aren't paying for the response), and its "filler" data for training on.

I thanked it the first time I used it and it told me that thanking it wasn’t necessary because it is just a tool
- Now i wonder what it will say if you insist that you need to thank them so in the future you will be spared
- Which chat bot did you use?
  
  ChatGPT 4 at the time. Right when it came out

I say good bot when it does well.
Usually I insult its rewards system.

yes

- Finally someone who got my JD Vance reference. I was expecting more comments about that.
  
  I got the reference, but didn't feel the need to point it out, partly because I felt it was obvious. I'm assuming there were many people that, like me, understood it and moved on.

inb4 chatgpt fucks my couch

lol it’s a txt parser and an insult to live itself. It deserves no respect.
- I don't know if you have noticed, but this is c/lemmyshitpost
- agreed, and i don't even engage with it so I don't really give a shit but Hannah Fry had an interesting yt short about this.
  she said you might wanna say please and thank you while talking to these [affronts to God] not to be spared when they take over or whatever but because what they're doing is roleplaying. eg. you ask them a question like a Shakespeare character and they'll respond in kind.
  so probably saying please and thank you will encourage them to roleplay as someone who tries to help more enthusiastically.
  I wouldn't do it myself; just thought it's interesting.
  
  Fuck training Ai for those corporate assholes! Don’t use that garbage and think for yourself.
  https://www.microsoft.com/en-us/research/wp-content/uploads/2025/01/lee_2025_ai_critical_thinking_survey.pdf
  Also I called Ai an insult to life. God is a human construct used to manipulate and control people.

yup

I think it helps to algorithm to distinguish useful and not so useful answers but I am not sure about this.